Dear Frode,
> It has been performed an activation study (PET) of two groups using the
> same study paradigm. There is one variable which in study group A is
> constant, and in study group B has two different levels (both higher
> values than in group A). I have been asked to see if this variable has
> a linear relationship across the three levels. Which approach is the
> most appropriate? Would it help (more valid?) to consider this as
> random effects, and apply a regression analysis afterwards?
Yes this seems very sensible. I assume you are looking for a condition
x 'level' interaction . This could be tested for with a fixed effect
analysis but is more simply implemented at a second level giving a
random effect analysis and generalization to the population at large.
You would simply create contrast images for each subject (using SPM99)
reflecting the condition effect and use these as new dependent
variables at the second level with your 'level' as a regressor.
I hope this helps - Karl
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|